Preservation of Privacy in Thwarting the Ballot Stuffing Scheme

Users of an online trading system rely on Reputation Systems to better judge whom should be trusted and to what degree. This is achieved through users building up reputations in the system. In these types of environments, it has been shown that users with good reputations do more business than users with bad reputations. The ballot stuffing scheme exploits this and has fraudulent users placing a number of false bids in an attempt to better the reputation of a single user. Though previous research has dealt with thwarting the one man ballot stuffing scheme, the issue of privacy was neglected. The solution proposed relied on looking up the coordinates of a user who is a cellular phone holder. Upon placing a bid, the user's geographical coordinates are compared to the coordinates of other users involved in the transaction. If the users were within a predefined distance to one another, the transaction was marked as suspicious. This mechanism relies on storing the coordinates of a user over time and, from a privacy perspective, is unacceptable. The intention of this paper is to propose several solutions that attempt to safeguard the privacy of all users involved when calculating the distance between two cellular phone holders, i.e., thwarting the one man ballot stuffing scheme. We discuss solutions that cater for service providers who may be willing or unwilling to participate in safeguarding the privacy of their users. These techniques include Secure Multi-party Computation, polynomial interpolation and the addition of untrusted third parties.